Method and device for constructing underground space flood emergency rescue and dispatch command model under heavy rain scenario

Abstract

The present invention discloses a method and device for constructing an emergency rescue and dispatch command model for underground space floods under heavy rain scenarios, which belongs to the field of construction safety management technology and solves the problem that the existing technology cannot ensure the effective implementation of emergency rescue plans. The present invention conducts flood simulation under multiple scenarios with different rainfall amounts based on heavy rain scenario data and a cellular automaton model, obtains flood simulation results to identify key risk influencing factors and conducts severity evaluation, and at the same time conducts risk level evaluation based on the underground space disaster risk evaluation index system under heavy rain scenarios, and matches the target underground space emergency rescue resource list with the constructed knowledge base to obtain a first emergency task list for application in the corresponding rainfall scenario, and uses a cellular automaton model for pre-deduction. According to the pre-deduction results, a second emergency resource list is generated, that is, an emergency rescue and dispatch command decision is obtained. The present invention is used for emergency rescue and dispatch command of underground space floods under heavy rain scenarios.

Description

Technical Field

[0001] A method and device for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm situation, which are used for underground space flood emergency rescue and dispatch command under a rainstorm situation and belong to the technical field of construction safety management. Background Art

[0002] In recent years, against the backdrop of global climate change, extreme weather events have become frequent. Urban flooding caused by heavy rain is a particularly prominent problem, posing a severe challenge to urban infrastructure. In densely populated metropolitan areas with extensive underground space development and utilization, underground space construction, as a crucial component of public infrastructure development, is directly related to public safety and the security of life and property. Therefore, research on emergency rescue and dispatch command methods for underground flooding in heavy rain scenarios has become a critical issue in urban management and public safety.

[0003] With the development of emergency management concepts and technologies, the application of modern information technology has become increasingly widespread, providing strong support for improving the speed and efficiency of emergency response. Governments and relevant departments are actively building an information-based emergency management system, leveraging technologies such as big data, cloud computing, the Internet of Things, and artificial intelligence to achieve accurate early warning, efficient response, and scientific decision-making for emergencies. For example, by building a model of the urban drainage system and integrating it with meteorological forecast data, it is possible to predict in advance where flooding may occur due to heavy rain. Utilizing a GIS (Geographic Information System), disaster-affected areas can be quickly located and rescue routes optimized. Intelligent monitoring and sensor networks can monitor underground water level fluctuations in real time, providing a basis for decision-making in rescue and dispatch. On this basis, the development and improvement of emergency response plans has become the core of emergency management. These emergency response plans detail the entire process, from warning information dissemination, emergency resource deployment, on-site response measures, to post-disaster reconstruction, and clearly define the responsibilities and coordination mechanisms of responsible parties at all levels. The implementation of these plans relies on an efficient dispatch and command system to ensure that emergency response can be rapidly initiated, resources can be rationally allocated, disasters can be effectively controlled, and losses minimized when heavy rain strikes.

[0004] However, the existing technology has the following technical problems:

[0005] 1. Existing disaster emergency rescue and dispatch command decision-making systems perform simple knowledge base statistics, matching the brief on-site situation with the emergency rescue plan. However, the plan lacks adaptability to the on-site situation and cannot guarantee the effective implementation of the emergency rescue plan.

[0006] 2. Research related to underground space flood emergency rescue and dispatch command simulation at home and abroad is concentrated on the operation period, and usually takes floods, personnel, and materials as the main research objects. Insufficient consideration is given to factors such as flood risk factors and construction equipment. Due to the lack of construction and simulation analysis of disaster scenarios of different levels, the emergency rescue plan is not intelligent enough, and the rescue effect cannot be quantitatively evaluated, making it difficult to guide the practice of underground space flood emergency rescue during the construction period. Summary of the Invention

[0007] The purpose of the present invention is to provide a method and device for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario, so as to solve the problem that the existing relevant disaster emergency rescue and dispatch command decision-making system performs simple knowledge base statistics, matches the brief situation on the scene with the emergency rescue plan, lacks the adaptability correction of the plan to the on-site situation, and cannot ensure the effective implementation of the emergency rescue plan; and the research related to underground space flood emergency rescue and dispatch command simulation at home and abroad is concentrated on the operation period, and usually takes floods, personnel, and materials as the main research objects, and does not give enough consideration to factors such as flood risk factors and construction equipment. Due to the lack of construction and simulation analysis of disaster scenarios of different levels, the emergency rescue plan is not intelligent enough, and the rescue effect cannot be quantitatively evaluated, making it difficult to guide the practice of flood emergency rescue during the construction period of underground space.

[0008] In order to achieve the above object, the technical solution adopted by the present invention is:

[0009] A method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario includes the following steps:

[0010] Step S1: Based on the rainstorm scenario data and the cellular automaton model, a flood simulation is performed under multiple scenarios with different rainfall amounts, the spatiotemporal evolution process of the waterlogging disaster is simulated and analyzed, and the flood simulation results are obtained, wherein the flood simulation results include the catchment area, water depth, arrival time, and inundation range;

[0011] Step S2: Based on the flood simulation results, identify key risk factors and conduct a severity assessment. Simultaneously, conduct a risk level assessment based on the underground space disaster risk assessment index system under a rainstorm scenario. Key risk factors include construction personnel, important facilities, protection plans, and underground space flooding scope.

[0012] Step S3: Matching the target underground space emergency rescue resource list obtained based on the severity evaluation and risk level assessment results and the statistically obtained list with the constructed knowledge base to obtain a multi-angle first emergency task list for the underground space, wherein the multi-angle first emergency task list for the underground space includes determining the types and quantities of emergency resources required for the flood accident emergency response based on the state of the flood accident;

[0013] Step S4: Apply the contents of the first emergency resource list to the corresponding rainfall scenario, and use the cellular automaton model for pre-simulation. Based on the pre-simulation results, generate the second emergency resource list, that is, obtain the emergency rescue and dispatch command decision.

[0014] Furthermore, the severity evaluation in step 2 includes the sum of the in-depth analysis and the disaster loss assessment;

[0015] The specific steps of the in-depth analysis include:

[0016] Step 2.1: Verify the key risk influencing factors IN(t)=(x c 0 (t)), t=1,2,3……M, whether S is satisfied c ={x c (1),x c (1),……,x c (M)},c=1,2,3 data format requirements, if the requirements are met, use the formula Perform normalization processing. If the requirements are not met, perform data verification and re-obtain input data for judgment. M represents the frame sequence of input data with time tags, IN and S are both set letters, and x c 0 (t) represents the input data before normalization;

[0017] Step 2.2: Construct a loss assessment model based on the normalized data obtained in Step 2.1. Specifically, assign descriptive values to each of the three disaster-causing factors: disaster variables, carrier variables, and control variables. The loss assessment model is constructed based on these three types of disaster-causing factors: disaster variables, including the extent of underground flooding; carrier variables, including the construction site, including construction personnel; and control variables, including protective facilities, including key facilities and protection plans.

[0018] The loss assessment model is expressed as:

[0019] x(t)=(x1(t),x2(t),x3(t)) T

[0020] IN:(x i 0 (t))(i=1,2,3)

[0021] OUT:(x i 0 (t+1))(i=1,2,3)

[0022] Among them, x1(t) represents the description value of a certain disaster variable of underground space flood at time t, x2(t) represents the description value of a certain carrier variable of underground station flood at time t, and x3(t) represents the description value of a certain control variable of underground station flood at time t. i 0 (t))(i=1,2,3) represents the format of input data of loss assessment model, OUT:(x i 0 (t+1))(i=1,2,3) represents the format of the output data of the loss assessment model, and x(t) represents the normalized input data;

[0023] Step 2.3: Based on the data set including the disaster factors, the loss assessment model is trained on the corresponding relationship between the disaster factors and the loss assessment. After the training, the predicted disaster factors are subjected to in-depth analysis and evaluation to obtain the in-depth analysis and evaluation value of the corresponding disaster factors.

[0024] The disaster loss assessment includes the sum of the risk population assessment and the economic loss assessment;

[0025] The risk population assessment refers to counting the total value of the risk population and obtaining the risk population assessment score corresponding to the total value according to the risk population assessment standard, wherein the risk population includes the casualties on site due to the disaster, the missing persons due to the disaster, the emergency relocated and resettled population, and the trapped population. The risk population assessment standard includes that if the total value of the risk population is greater than or equal to 100 people, the corresponding risk population assessment score is 5 points; if the total value of the risk population is greater than or equal to 50 people and less than 100 people, the corresponding risk population assessment score is 4 points; if the total value of the risk population is greater than or equal to 10 people and less than 50 people, the corresponding risk population assessment score is 3 points; if the total value of the risk population is greater than or equal to 3 people and less than 10 people, the corresponding risk population assessment score is 2 points; if the total value of the risk population is less than 3 people, the corresponding risk population assessment score is 1 point;

[0026] The economic loss evaluation refers to calculating the total economic loss and obtaining the economic loss evaluation score corresponding to the total economic loss according to the economic loss evaluation standard, wherein the total economic loss includes the economic losses of construction site equipment loss, construction period loss and material loss at the construction site after the flood occurs. The economic loss evaluation standard includes that the total economic loss is greater than or equal to 1 million yuan, and the corresponding economic loss evaluation score is 5 points, indicating that particularly serious economic losses are caused; the total economic loss is greater than or equal to 500,000 yuan and less than 1 million yuan, and the corresponding economic loss evaluation score is 4 points, indicating that serious economic losses are caused; the total economic loss is greater than or equal to 100,000 yuan and less than 500,000 yuan, and the corresponding economic loss evaluation score is 3 points, indicating that relatively heavy economic losses are caused; the total economic loss is greater than or equal to 10,000 yuan and less than 100,000 yuan, and the corresponding economic loss evaluation score is 2 points, indicating that certain economic losses are caused; the total economic loss is less than 10,000 yuan, and the corresponding economic loss evaluation score is 1 point, indicating that basically no economic losses are caused.

[0027] Furthermore, the specific steps of the risk level assessment in step 2 are:

[0028] Construct a multi-index comprehensive evaluation system, that is, construct an underground space disaster risk assessment index system under heavy rain scenarios, which includes overall indicators, primary indicators, and secondary indicators. Among them, the overall indicator is the underground space flood risk under heavy rain scenarios. The primary indicators include the hazard of disaster-causing factors, the sensitivity of the disaster-prone environment, the vulnerability of the disaster-bearing body, and the disaster prevention and mitigation capabilities. The secondary indicators include the 24-hour maximum rainfall, rainfall duration, construction management and responsibility implementation for the hazard of disaster-causing factors; the catchment area, drainage network density, complex surrounding environment, and rivers for the sensitivity of the disaster-prone environment; the drainage ditch drainage capacity and foundation pit depth for the vulnerability of the disaster-bearing body; the height of the retaining wall, disaster monitoring and early warning technology, and emergency material reserves for the disaster prevention and mitigation capabilities;

[0029] A grey clustering evaluation method is constructed to score the secondary indicators in the multi-indicator comprehensive evaluation system to determine the primary indicator score. Finally, the risk level assessment of the total indicator is obtained based on the weight of the given primary indicator. The specific steps are as follows:

[0030] Step 2-1: Establish a whitening weight function for the quantitative risk assessment of rainstorm floods around underground spaces, namely the grey clustering evaluation method. Utilize the whitening weight function in combination with the quantitative risk characteristics of rainstorm floods around underground spaces to divide the risk level into five grey classes. Based on the threshold center point vector U = (9, 7, 5, 3, 1), the reference threshold and corresponding risk level of each grey class are obtained. Based on the reference threshold and risk level of each grey class, the weight w of all secondary indicators under the i-th primary indicator is assigned. i =(w i1 ,w i2 ,w i3 ,...wij ), among which, the quantitative risk characteristics of rainstorm flooding around underground space are the secondary indicators in the multi-index comprehensive evaluation system. When the gray category is 1, the reference threshold is [0, 9, ∞], and the risk level is very safe. When the gray category is 2, the reference threshold is [0, 7, 14], and the risk level is safe. When the gray category is 3, the reference threshold is [0, 5, 10], and the risk level is general. When the gray category is 4, the reference threshold is [0, 3, 6], and the risk level is dangerous. When the gray category is 5, the reference threshold is [0, 1, 2], and the risk level is very dangerous.

[0031] The formula of the whitening weight function is:

[0032]

[0033]

[0034]

[0035]

[0036]

[0037] Where, d ijk represents the score of expert k on the jth secondary indicator under the i-th primary indicator, k = 1, 2, 3, ..., f1 1 (d ijk )、f1 2 (d ijk )、f1 3 (d ijk )、f1 4 (d ijk )、f1 5 (d ijk ) represent the risk levels of the jth secondary indicator under the i-th primary indicator in each gray category;

[0038] Step 2-2: Construct a secondary grey evaluation matrix. The specific steps are as follows:

[0039] By d ijk Obtain the jth secondary index c under the i-th primary index ij The evaluation coefficient in each gray class is as follows:

[0040]

[0041] Calculate the secondary index c separately ij The evaluation coefficient in the five gray categories is as follows:

[0042]

[0043] Among them, e represents the e-th gray class, Y ij represents the total clustering coefficient,

[0044] Based on the total clustering coefficient Y ij and evaluation coefficient Y ije Get the gray evaluation weight value:

[0045]

[0046] Arrange the grey evaluation weights of all secondary indicators under the first-level indicator of each grey category and establish the secondary grey evaluation matrix R i :

[0047]

[0048] Step 2-3: Calculate the comprehensive evaluation results, that is, the weights w of all secondary indicators under the i-th primary indicator i =(w i1 ,w i2 ,w i3 ,...w ij ) and the secondary grey evaluation matrix R i Multiply them to get the weighted evaluation result Z i , that is, Z i =W i R i A is the first-level indicator i The risk level assessment results, similarly, according to the first-level indicator A i The first-level index weight W is constructed based on the risk level assessment results, and the first-level grey evaluation matrix Z is constructed to obtain the risk level assessment result M=WZ of the total index of the rainstorm flood around the underground space. * =M·UT converts the first-level matrix M into a comprehensive evaluation value W * and through the comprehensive evaluation value W * The reference quantitative value range that falls into determines the risk level of the target, where ·UT represents performing a transpose combination transformation.

[0049] Furthermore, the specific steps of step S4 are:

[0050] Step S4.1: Establish a hierarchical emergency decision-making mechanism based on leadership decision-making, professional handling, and public response;

[0051] Step S4.2: Based on the professional handling in the emergency decision-making mechanism and the first emergency rescue resource list, the contents of the first emergency resource list are applied to the corresponding rainfall scenario, and a cellular automaton model is used for pre-simulation. According to the pre-simulation results, a second emergency resource list is generated, that is, an emergency rescue and dispatch command decision is obtained, wherein the emergency rescue and dispatch decision includes generating personnel evacuation paths and generating material utilization directions.

[0052] Furthermore, in step S4.2,

[0053] The specific steps of generating a disaster evacuation path for personnel are:

[0054] Step 4.21: Obtain basic information about the underground space and construct a digital model of the underground space using 3D modeling software. The basic information includes CAD drawings of the underground space, 3D scanning data, camera distribution, and access control system information.

[0055] Step 4.22: Conduct disaster simulation based on the pre-simulation results of the underground space to determine all possible evacuation paths in the underground space under the current scenario;

[0056] Step 4.23: Determine the optimal disaster avoidance path based on the determined underground disaster avoidance paths under the current situation and the Dijkstra algorithm; the specific steps are:

[0057] Step 4.231. Create a directed weighted graph G(V, E, W), where V is used to store all nodes in the underground space disaster avoidance path under the current scenario, E is used to store the edges connecting two nodes in the underground space disaster avoidance path under the current scenario, and W is used to store all navigable paths in the underground space disaster avoidance path under the current scenario, i.e., the disaster avoidance path.

[0058] Step 4.232: Define a set D to store the emergency evacuation nodes in the underground space disaster avoidance path under the current situation;

[0059] Step 4.233. Create a set P to store the sequence of all nodes in the directed weighted graph G(V, E, W), and create a set Q to record all node paths connected to each node in the directed weighted graph G(V, E, W).

[0060] Step 4.234: Based on each emergency evacuation node in set D, use the Dijkstar algorithm to perform path resolution on sets P and Q, obtain a shortest path, and simultaneously construct sets M and N. The points and edges on the shortest path are traversed and stored in sets M and N respectively.

[0061] Step 4.235: For each emergency evacuation node, remove the edge of the shortest path stored in the set N in step 4.234 from the corresponding directed weighted graph G, so that the directed weighted graph G becomes a directed weighted graph G1;

[0062] Step 4.236: If there is no traversable path from each emergency evacuation node to the target node in the directed weighted graph G1, proceed to step 42.37. Otherwise, proceed to step 4.233 and continue intermittently based on the directed weighted graph G1 until all shortest paths from each emergency evacuation node to the target node or the first z shortest paths are obtained, and the loop ends. If all shortest paths are less than z, all shortest paths from each emergency evacuation node to the target node are obtained. Otherwise, the first z shortest paths are obtained.

[0063] Step 42.37: Given a threshold target, if all shortest paths or the shortest paths of the first z paths are less than or equal to the threshold target, all shortest paths are considered the optimal disaster avoidance paths. Otherwise, a screening function is used to determine the optimal disaster avoidance paths with a number of paths equal to the threshold target among all shortest paths or the shortest paths of the first z paths.

[0064] Step 4.24: During the disaster avoidance process, re-execute step 4.21 based on the real-time monitoring of rainfall conditions and changes in the underground space;

[0065] The material utilization direction is generated by using a CNN neural network model and pre-deduction results. The CNN neural network model includes two objective functions and five constraints that need to be met.

[0066] Objective function 1 is used to represent the lowest cost of emergency material dispatch:

[0067]

[0068] Among them, m represents the number of emergency material rescue points, n represents the number of disaster-affected points, x(Z v ) qw Indicates the emergency supplies from point w to rescue point R w Towards the qth disaster point D q Dispatch type v emergency supplies Z v The number of G qw Indicates the emergency supplies from point w to rescue point R w Towards the qth disaster point D q The number of times the vth type of material is dispatched;

[0069] Objective function 2 is used to express the shortest adjustment time of emergency supplies;

[0070]

[0071] in, Indicates that the emergency supplies at point w are sent to rescue point R w Towards the qth disaster point D q Triangular fuzzy number of dispatching time for allocating emergency supplies, t wq1 is the pessimistic value of the fuzzy number, t wq2 is the normal value of the fuzzy number, t wq3 is the optimistic value of the fuzzy number, Indicates the emergency supplies from point w to rescue point R w Whether to deploy supplies to the qth disaster site D q ,

[0072] There are five constraints:

[0073] Constraint 1 indicates that the actual total supply of emergency supplies at each emergency rescue point is equal to the total demand for emergency supplies at each disaster site. The formula is:

[0074]

[0075] Among them, r(Z v ) w Indicates the w-th emergency supplies rescue point R w The vth material Z v The actual supply, d~(Z v ) q The emergency supplies Z at the qth disaster site v The actual demand;

[0076] Constraint 2 indicates that the vth type of emergency supplies Z is actually dispatched from each emergency supply rescue point to the qth disaster point. v The quantity is equal to the emergency supplies Z of the disaster site v The demand is:

[0077]

[0078] Among them, x(Z v ) wq The emergency supplies Z actually dispatched from the w-th emergency supply rescue point to the q-th disaster site v the number of represents the demand for emergency supplies at the qth disaster site;

[0079] Constraint 3 indicates that the emergency supplies Z are allocated from the w-th emergency supply rescue point to each disaster site. v The sum of the quantity of emergency supplies is equal to the emergency supplies transported from the rescue point Z v The actual supply is:

[0080]

[0081] Constraint 4 indicates that the emergency supplies from point w are sent to rescue point R. w Deployed to the qth disaster site D q The quantity of emergency supplies is a non-negative number, and the formula is:

[0082] x(Z v ) wq ≥0

[0083] Constraint 5 indicates that the emergency supplies from point w are sent to rescue point R. w Whether to deploy emergency supplies to the qth disaster site D q If there are emergency supplies from the wth emergency supplies rescue point R w Deployed to the qth disaster site D q , then x~ wq The value of is 1, otherwise it is 0;

[0084]

[0085] Here, k represents the number of species.

[0086] Furthermore, the cellular automaton model simulation in steps S1 and S4 is based on the rainstorm scenario data, the underground space construction model and the catchment area, using the cellular automaton model to perform physical simulation and numerical simulation of the disaster evolution to obtain disaster data, including the underground space flooded area, flooded water depth and flood arrival time;

[0087] Among them, the underground space construction model is obtained by obtaining underground space data and importing it based on the underground space data after remodeling according to the drawings. The underground space data includes DEM data, terrain model, remote sensing image, BIM model, rainstorm data and meteorological data. The terrain model refers to the underground space model. The rainstorm scenario data and the catchment area are obtained based on the rainfall parameter information in the underground space data and the obtained rainstorm intensity.

[0088] A device for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario includes a memory, a processor and a computer program stored in the memory. The processor executes the computer program to implement the steps of a method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario.

[0089] Compared with the prior art, the advantages of the present invention are:

[0090] The present invention uses a flood simulation model to conduct a basic assessment of the disaster situation, then conducts a severity evaluation and risk level assessment, and generates a first emergency resource list. The first emergency resource list is then pre-simulated to generate a second emergency resource list as the final basis for on-site emergency resource allocation and scheduling simulation. The relevant data is simultaneously stored in a knowledge base to provide a theoretical basis for subsequent flood emergency management. By analyzing the rationality of the on-site human, machine, and material settings, the impact of flood risks is minimized. By generating personnel evacuation routes and generating material utilization directions, real-time response and plan adjustments to flood emergency rescue commands are achieved, further reducing on-site disaster losses. This is specifically reflected in the following aspects:

[0091] First, the present invention uses a flood simulation model to conduct basic disaster analysis, then conduct severity assessment and risk level evaluation, and generate a first emergency resource list. This list is then pre-simulated to generate a second emergency resource list, which serves as the final basis for on-site emergency resource allocation and scheduling simulation. This allows for on-site adaptation corrections to ensure the effective implementation of emergency rescue plans.

[0092] 2. The present invention fully considers factors such as flood risk factors and construction equipment, and can construct and simulate analyses for disaster scenarios of different levels. The emergency rescue plan is highly intelligent, so that the rescue effect can be quantitatively evaluated. That is, through severity evaluation and risk level evaluation, a method for quantitative evaluation of underground space disasters under heavy rain scenarios is given, and a flood risk level evaluation mechanism is established, which provides a quantitative basis for emergency rescue and dispatch command of heavy rain and floods under different scenarios, so as to effectively guide the practice of emergency rescue of floods during the construction period of underground space. BRIEF DESCRIPTION OF THE DRAWINGS

[0093] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other relevant drawings can be obtained based on these drawings without paying any creative work.

[0094] Figure 1 It is a basic logic diagram of the present invention;

[0095] Figure 2 It is a flowchart of the application of severity evaluation and risk level assessment results;

[0096] Figure 3 It is a flow chart generated by the emergency rescue and dispatch command plan. DETAILED DESCRIPTION

[0097] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0098] A method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario, characterized by comprising the following steps:

[0099] Step S1, based on the rainstorm scenario data and the cellular automaton model, conduct flood simulation under multiple scenarios with different rainfall amounts, simulate and analyze the spatiotemporal evolution process of waterlogging disasters, and obtain flood simulation results, wherein the flood simulation results include the catchment area, water depth, arrival time and flooding range; the cellular automaton model simulation is based on the rainstorm scenario data, the underground space construction model and the catchment area, and uses the cellular automaton model to perform physical simulation and numerical simulation of the disaster evolution to obtain disaster data, including the underground space flooding area, flooding depth and flood arrival time; wherein, the underground space construction model is obtained by obtaining underground space data, and importing it based on the underground space data after reproducing the model according to the drawing. The underground space data includes DEM data, terrain model, remote sensing image, BIM model, rainstorm data and meteorological data. The terrain model refers to the underground space model. The rainstorm scenario data and the catchment area are obtained based on the rainfall parameter information in the underground space data and the obtained rainstorm intensity.

[0100] Step S2: Based on the flood simulation results, identify key risk factors and conduct a severity assessment. Simultaneously, conduct a risk level assessment based on the underground space disaster risk assessment index system under a rainstorm scenario. Key risk factors include construction personnel, important facilities, protection plans, and underground space flooding scope.

[0101] Severity evaluation includes the sum of in-depth analysis and disaster loss assessment;

[0102] The specific steps of the in-depth analysis include:

[0103] Step 2.1: Verify the key risk influencing factors IN(t)=(x c 0 (t)), t=1,2,3……M, whether S is satisfied c ={x c (1),x c (1),……,x c (M)},c=1,2,3 data format requirements, if the requirements are met, use the formula Perform normalization processing. If the requirements are not met, perform data verification and re-obtain input data for judgment. M represents the frame sequence of input data with time tags, IN and S are both set letters, and x c 0 (t) represents the input data before normalization;

[0104] Step 2.2: Construct a loss assessment model based on the normalized data obtained in Step 2.1. Specifically, assign descriptive values to each of the three disaster-causing factors: disaster variables, carrier variables, and control variables. The loss assessment model is constructed based on these three types of disaster-causing factors: disaster variables, including the extent of underground flooding; carrier variables, including the construction site, including construction personnel; and control variables, including protective facilities, including key facilities and protection plans.

[0105] The loss assessment model is expressed as:

[0106] x(t)=(x1(t),x2(t),x3(t)) T

[0107] IN:(x i 0 (t))(i=1,2,3)

[0108] OUT:(x i 0 (t+1))(i=1,2,3)

[0109] Among them, x1(t) represents the description value of a certain disaster variable of underground space flood at time t, x2(t) represents the description value of a certain carrier variable of underground station flood at time t, and x3(t) represents the description value of a certain control variable of underground station flood at time t. i 0 (t))(i=1,2,3) represents the format of input data of loss assessment model, OUT:(x i 0 (t+1))(i=1,2,3) represents the format of the output data of the loss assessment model, and x(t) represents the normalized input data;

[0110] Step 2.3: Based on the data set including the disaster-causing factors, the loss assessment model is trained on the correspondence between the disaster-causing factors and the loss assessment. After the training, the predicted disaster-causing factors are deeply analyzed and evaluated to obtain the deep analysis evaluation value of the corresponding disaster-causing factors.

[0111] The disaster loss assessment includes the sum of the risk population assessment and the economic loss assessment;

[0112] The risk population assessment refers to counting the total value of the risk population and obtaining the risk population assessment score corresponding to the total value according to the risk population assessment standard, wherein the risk population includes the casualties on site due to the disaster, the missing persons due to the disaster, the emergency relocated and resettled population, and the trapped population. The risk population assessment standard includes that if the total value of the risk population is greater than or equal to 100 people, the corresponding risk population assessment value is 5 points; if the total value of the risk population is greater than or equal to 50 people and less than 100 people, the corresponding risk population assessment value is 4 points; if the total value of the risk population is greater than or equal to 10 people and less than 50 people, the corresponding risk population assessment value is 3 points; if the total value of the risk population is greater than or equal to 3 people and less than 10 people, the corresponding risk population assessment value is 2 points; if the total value of the risk population is less than 3 people, the corresponding risk population assessment value is 1 point, as shown in Table 1:

[0113] Table 1 Risk population assessment table

[0114] Number of people at risk Corresponding score More than 100 people 5 More than or equal to 50 people and less than 100 people 4 More than or equal to 10 people and less than 50 people 3 More than or equal to 3 people and less than 10 people 2 Less than 3 people 1

[0115] The economic loss evaluation refers to calculating the total economic loss and obtaining the economic loss evaluation score corresponding to the total economic loss according to the economic loss evaluation standard, wherein the total economic loss includes the economic losses of construction site equipment loss, construction period loss and material loss at the construction site after the flood occurs. The economic loss evaluation standard includes that the total economic loss is greater than or equal to 1 million yuan, and the corresponding economic loss evaluation score is 5 points, indicating that extremely serious economic losses are caused; the total economic loss is greater than or equal to 500,000 yuan and less than 1 million yuan, and the corresponding economic loss evaluation score is 4 points, indicating that serious economic losses are caused; the total economic loss is greater than or equal to 100,000 yuan and less than 500,000 yuan, and the corresponding economic loss evaluation score is 3 points, indicating that relatively heavy economic losses are caused; the total economic loss is greater than or equal to 10,000 yuan and less than 100,000 yuan, and the corresponding economic loss evaluation score is 2 points, indicating that certain economic losses are caused; the total economic loss is less than 10,000 yuan, and the corresponding economic loss evaluation score is 1 point, indicating that basically no economic losses are caused, as shown in Table 2:

[0116] Table 2 Economic loss evaluation table

[0117] Impact of economic losses Evaluation Criteria Corresponding score Causing particularly serious economic losses Greater than or equal to 1 million yuan 5 Cause serious economic losses Greater than or equal to 500,000 yuan and less than 1 million yuan 4 Caused heavy economic losses Greater than or equal to 100,000 yuan, less than 500,000 yuan 3 Cause certain economic losses Greater than or equal to 10,000 yuan, less than 100,000 yuan 2 Basically no economic losses were caused Less than 10,000 yuan 1

[0118] The specific steps of risk level assessment are:

[0119] A multi-index comprehensive evaluation system is constructed, that is, an underground space disaster risk assessment index system under heavy rain scenarios is constructed, which includes a total index, a primary index, and a secondary index. Among them, the total index is the underground space flood risk under heavy rain scenarios, the primary index includes the hazard of disaster-causing factors, the sensitivity of the disaster-prone environment, the vulnerability of the disaster-bearing body, and the disaster prevention and mitigation capabilities, the secondary index includes the 24-hour maximum rainfall, rainfall duration, and construction management and responsibility implementation for the hazard of disaster-causing factors, the catchment area, drainage network density, complex surrounding environment, and rivers for the sensitivity of the disaster-prone environment, the drainage capacity of the drainage ditch and the depth of the foundation pit for the vulnerability of the disaster-bearing body, and the height of the retaining wall, disaster monitoring and early warning technology, and emergency material reserves for the disaster prevention and mitigation capabilities, as shown in Table 3. The total index is represented by A, the primary index is represented by B, and the secondary index is represented by C:

[0120] Table 3. Deep foundation pit engineering disaster risk assessment index system under heavy rain scenario

[0121]

[0122]

[0123] A grey clustering evaluation method is constructed to score the secondary indicators in the multi-indicator comprehensive evaluation system to determine the primary indicator score. Finally, the risk level assessment of the total indicator is obtained based on the weight of the given primary indicator. The specific steps are as follows:

[0124] Step 2-1: Establish a whitening weight function for the quantitative risk assessment of rainstorm floods around underground spaces, namely the grey clustering evaluation method. Utilize the whitening weight function in combination with the quantitative risk characteristics of rainstorm floods around underground spaces to divide the risk level into five grey classes. Based on the threshold center point vector U = (9, 7, 5, 3, 1), the reference threshold and corresponding risk level of each grey class are obtained. Based on the reference threshold and risk level of each grey class, the weight w of all secondary indicators under the i-th primary indicator is assigned. i =(w i1 ,w i2 ,w i3 ,...w ij ), among which, the quantitative risk characteristics of rainstorm flooding around underground space are the secondary indicators in the multi-index comprehensive evaluation system. When the gray category is 1, the reference threshold is [0, 9, ∞], and the risk level is very safe. When the gray category is 2, the reference threshold is [0, 7, 14], and the risk level is safe. When the gray category is 3, the reference threshold is [0, 5, 10], and the risk level is general. When the gray category is 4, the reference threshold is [0, 3, 6], and the risk level is dangerous. When the gray category is 5, the reference threshold is [0, 1, 2], and the risk level is very dangerous, as shown in Table 4:

[0125] Table 4 Reference thresholds and risk levels for each gray category

[0126]

[0127]

[0128] The formula of the whitening weight function is:

[0129]

[0130]

[0131]

[0132]

[0133]

[0134] Where, d ijk represents the score of expert k on the jth secondary indicator under the i-th primary indicator, k = 1, 2, 3, ..., f1 1 (d ijk )、f1 2 (d ijk )、f1 3 (d ijk )、f1 4 (d ijk )、f1 5 (d ijk ) represent the risk levels of the jth secondary indicator under the i-th primary indicator in each gray category;

[0135] Step 2-2: Construct a secondary grey evaluation matrix. The specific steps are as follows:

[0136] By d ijk Obtain the jth secondary index c under the i-th primary index ij The evaluation coefficient in each gray class is as follows:

[0137]

[0138] Calculate the secondary index c separately ij The evaluation coefficient in the five gray categories is as follows:

[0139]

[0140] Among them, e represents the e-th gray class, Y ij represents the total clustering coefficient,

[0141] Based on the total clustering coefficient Y ij and evaluation coefficient Y ije Get the gray evaluation weight value:

[0142]

[0143] Arrange the grey evaluation weights of all secondary indicators under the first-level indicator of each grey category and establish the secondary grey evaluation matrix R i :

[0144]

[0145] Step 2-3: Calculate the comprehensive evaluation results, that is, the weights w of all secondary indicators under the i-th primary indicator i =(w i1 ,w i2 ,w i3 ,...w ij ) and the secondary grey evaluation matrix R i Multiply them to get the weighted evaluation result Z i , that is, Z i =W i R i A is the first-level indicator i The risk level assessment results, similarly, according to the first-level indicator A i The first-level index weight W is constructed based on the risk level assessment results, and the first-level grey evaluation matrix Z is constructed to obtain the risk level assessment result M=WZ of the total index of the rainstorm flood around the underground space. * =M·UT converts the first-level matrix M into a comprehensive evaluation value W * and through the comprehensive evaluation value W * The reference quantitative value range that falls into determines the risk level of the target, where ·UT represents performing a transpose combination transformation.

[0146] Step S3: Based on the severity evaluation and risk level assessment results and the statistically obtained target underground space emergency rescue resource list, the constructed knowledge base is matched to obtain a multi-angle first emergency task list for the underground space, wherein the multi-angle first emergency task list for the underground space includes determining the types and quantities of emergency resources required for the emergency response to the flood accident according to the state of the flood accident; Figure 2 As shown, Figure 2 This is a flowchart of the application of severity evaluation and risk level assessment results provided by an embodiment of the present application. By sorting out flood emergency rescue and dispatch command plans under historical rainstorm scenarios, an emergency rescue knowledge base can be formed. Simultaneously, combined with the on-site materials, personnel, equipment, and other related conditions of the target underground space, an emergency rescue resource list can be formed. Based on the target underground space emergency rescue resource list and the constructed knowledge base, combined with the simulation of the on-site disaster situation, a first emergency task list that matches the level of the rainstorm disaster can be obtained.

[0147] Step S4: Apply the contents of the first emergency resource list to the corresponding rainfall scenario and perform a pre-simulation using a cellular automaton model (same simulation step as step S1). Based on the pre-simulation results, a second emergency resource list is generated, which is used to obtain the emergency rescue and dispatch command decision. The specific steps are:

[0148] Step S4.1: Establish a hierarchical emergency decision-making mechanism based on leadership decision-making, professional handling, and public response;

[0149] Step S4.2: Based on the professional handling in the emergency decision-making mechanism and the first emergency rescue resource list, the contents of the first emergency resource list are applied to the corresponding rainfall scenario, and a cellular automaton model is used for pre-simulation. According to the pre-simulation results, a second emergency resource list is generated, that is, an emergency rescue and dispatch command decision is obtained, wherein the emergency rescue and dispatch decision includes generating personnel evacuation paths and generating material utilization directions.

[0150] In step S4.2, the specific steps of generating a disaster evacuation path for personnel are:

[0151] Step 4.21: Obtain basic information about the underground space and construct a digital model of the underground space using 3D modeling software. The basic information includes CAD drawings of the underground space, 3D scanning data, camera distribution, and access control system information.

[0152] Step 4.22: Conduct disaster simulation based on the pre-simulation results of the underground space to determine all possible evacuation paths in the underground space under the current scenario;

[0153] Step 4.23: Determine the optimal disaster avoidance path based on the determined underground disaster avoidance paths under the current situation and the Dijkstra algorithm; the specific steps are:

[0154] Step 4.231. Create a directed weighted graph G(V, E, W), where V is used to store all nodes in the underground space disaster avoidance path under the current scenario, E is used to store the edges connecting two nodes in the underground space disaster avoidance path under the current scenario, and W is used to store all navigable paths in the underground space disaster avoidance path under the current scenario, i.e., the disaster avoidance path.

[0155] Step 4.232: Define a set D to store the emergency evacuation nodes in the underground space disaster avoidance path under the current situation;

[0156] Step 4.233. Create a set P to store the sequence of all nodes in the directed weighted graph G(V, E, W), and create a set Q to record all node paths connected to each node in the directed weighted graph G(V, E, W).

[0157] Step 4.234: Based on each emergency evacuation node in set D, use the Dijkstar algorithm to perform path resolution on sets P and Q, obtain a shortest path, and simultaneously construct sets M and N. The points and edges on the shortest path are traversed and stored in sets M and N respectively.

[0158] Step 4.235: For each emergency evacuation node, remove the edge of the shortest path stored in the set N in step 4.234 from the corresponding directed weighted graph G, so that the directed weighted graph G becomes a directed weighted graph G1;

[0159] Step 4.236: If there is no traversable path from each emergency evacuation node to the target node in the directed weighted graph G1, proceed to step 42.37. Otherwise, proceed to step 4.233 and continue intermittently based on the directed weighted graph G1 until all shortest paths from each emergency evacuation node to the target node or the first z shortest paths are obtained, and the loop ends. If all shortest paths are less than z, all shortest paths from each emergency evacuation node to the target node are obtained. Otherwise, the first z shortest paths are obtained.

[0160] Step 42.37: Given a threshold target, if all shortest paths or the shortest paths of the first z paths are less than or equal to the threshold target, all shortest paths are considered the optimal disaster avoidance paths. Otherwise, a screening function is used to determine the optimal disaster avoidance paths with a number of paths equal to the threshold target among all shortest paths or the shortest paths of the first z paths.

[0161] Step 4.24: During the disaster avoidance process, re-execute step 4.21 based on the real-time monitoring of rainfall conditions and changes in the underground space;

[0162] The material utilization direction is generated by using a CNN neural network model and pre-deduction results. The CNN neural network model includes two objective functions and five constraints that need to be met.

[0163] Objective function 1 is used to represent the lowest cost of emergency material dispatch:

[0164]

[0165] Among them, m represents the number of emergency material rescue points, n represents the number of disaster-affected points, x(Z v ) qw Indicates the emergency supplies from point w to rescue point R w Towards the qth disaster point D q Dispatch type v emergency supplies Z v The number of G qw Indicates the emergency supplies from point w to rescue point R w Towards the qth disaster point D q The number of times the vth type of material is dispatched;

[0166] Objective function 2 is used to express the shortest adjustment time of emergency supplies;

[0167]

[0168] in, Indicates that the emergency supplies at point w are sent to rescue point R w Towards the qth disaster point D q Triangular fuzzy number of dispatching time for allocating emergency supplies, t wq1 is the pessimistic value of the fuzzy number, t wq2 is the normal value of the fuzzy number, t wq3 is the optimistic value of the fuzzy number, Indicates the emergency supplies from point w to rescue point R w Whether to deploy supplies to the qth disaster site D q ,

[0169] There are five constraints:

[0170] Constraint 1 indicates that the actual total supply of emergency supplies at each emergency rescue point is equal to the total demand for emergency supplies at each disaster site. The formula is:

[0171]

[0172] Among them, r(Z v ) w Indicates the w-th emergency supplies rescue point R w The vth material Z v The actual supply, d~(Z v ) q The emergency supplies Z at the qth disaster site v The actual demand;

[0173] Constraint 2 indicates that the vth type of emergency supplies Z is actually dispatched from each emergency supply rescue point to the qth disaster point. v The quantity is equal to the emergency supplies Z of the disaster site v The demand is:

[0174]

[0175] Among them, x(Z v ) wq The emergency supplies Z actually dispatched from the w-th emergency supply rescue point to the q-th disaster site v the number of represents the demand for emergency supplies at the qth disaster site;

[0176] Constraint 3 indicates that the emergency supplies Z are allocated from the w-th emergency supply rescue point to each disaster site. v The sum of the quantity of emergency supplies is equal to the emergency supplies transported from the rescue point Z v The actual supply is:

[0177]

[0178] Constraint 4 indicates that the emergency supplies from point w are sent to rescue point R. w Deployed to the qth disaster site D q The quantity of emergency supplies is a non-negative number, and the formula is:

[0179] x(Z v ) wq ≥0

[0180] Constraint 5 indicates that the emergency supplies from point w are sent to rescue point R. w Whether to deploy emergency supplies to the qth disaster site D q If there are emergency supplies from the wth emergency supplies rescue point R w Deployed to the qth disaster site D q ,but The value of is 1, otherwise it is 0;

[0181]

[0182] Here, k represents the number of species.

[0183] like Figure 3 As shown, Figure 3 It is a schematic diagram of a dispatching and commanding decision output provided by an embodiment of the present application. By applying the contents of the first emergency resource list to the corresponding rainfall scenario and using the cellular automaton model for pre-deduction, the first emergency resource can be verified. Here, if the verification directly meets the requirements, the first emergency resource list is converted into a second emergency resource list for outputting emergency rescue and dispatching and command decisions. If it does not meet the requirements, it is necessary to re-acquire the flood-related information, and in combination with the deduction results, adjust the type and quantity information of emergency resources for the emergency response to the flood accident, and use the cellular automaton model for secondary deduction to make the emergency resource list meet the requirements, thereby generating a second emergency resource list.

Claims

1. A method for constructing an underground space flood emergency rescue and dispatch command model under heavy rain conditions, characterized in that: The process includes the following steps: Step S1: Based on the rainstorm scenario data and the cellular automaton model, a flood simulation is performed under multiple scenarios with different rainfall amounts, the spatiotemporal evolution process of the waterlogging disaster is simulated and analyzed, and the flood simulation results are obtained, wherein the flood simulation results include the catchment area, water depth, arrival time, and inundation range; Step S2: Based on the flood simulation results, identify key risk factors and conduct a severity assessment. Simultaneously, conduct a risk level assessment based on the underground space disaster risk assessment index system under a rainstorm scenario. Key risk factors include construction personnel, important facilities, protection plans, and underground space flooding scope. Severity evaluation includes the sum of in-depth analysis and disaster loss assessment; The specific steps of the in-depth analysis include: Step 2.1: Verify whether the key risk influencing factors used as input data meet the data format requirements. If they do, perform normalization processing. If they do not, perform data verification and re-obtain input data for judgment; Step 2.2: Construct a loss assessment model based on the normalized data obtained in Step 2.

1. Specifically, assign descriptive values to each of the three disaster-causing factors: disaster variables, carrier variables, and control variables. The loss assessment model is constructed based on these three types of disaster-causing factors: disaster variables, including the extent of underground flooding; carrier variables, including the construction site, including construction personnel; and control variables, including protective facilities, including key facilities and protection plans. Step 2.3: Based on the data set including the disaster factors, the loss assessment model is trained on the corresponding relationship between the disaster factors and the loss assessment. After the training, the predicted disaster factors are subjected to in-depth analysis and evaluation to obtain the in-depth analysis and evaluation value of the corresponding disaster factors. The disaster loss assessment includes the sum of the risk population assessment and the economic loss assessment; The risk population assessment refers to calculating the total value of the risk population and obtaining the risk population assessment score corresponding to the total value according to the risk population assessment standard; The economic loss assessment refers to calculating the total economic loss and obtaining the economic loss assessment score corresponding to the total economic loss according to the economic loss assessment standard; The specific steps of risk level assessment are: Construct a multi-index comprehensive evaluation system, that is, construct an underground space disaster risk evaluation index system under heavy rain scenarios including overall indicators, primary indicators and secondary indicators; A grey clustering evaluation method is constructed to score the secondary indicators in the multi-indicator comprehensive evaluation system to determine the primary indicator score. Finally, the risk level assessment of the total indicator is obtained based on the weight of the given primary indicator. Specifically: Calculate the comprehensive evaluation results, that is, The weights of all secondary indicators under the first-level indicator and the secondary grey evaluation matrix Multiply them to get the weighted evaluation result ,Right now First-level indicator The risk level assessment results, similarly, according to the first-level indicators The first-level indicator weights are constructed based on the risk level assessment results , and construct a first-level grey evaluation matrix Then we can get the risk level assessment result of the total quantitative index of rainstorm flood around underground space. , through the formula The first-level matrix Converted into comprehensive evaluation value and through comprehensive evaluation value The reference quantitative value range falls within to determine the risk level of the target, where Indicates the execution of transposition combination transformation; Step S3: Matching the target underground space emergency rescue resource list obtained based on the severity evaluation and risk level assessment results and the statistically obtained list with the constructed knowledge base to obtain a multi-angle first emergency task list for the underground space, wherein the multi-angle first emergency task list for the underground space includes determining the types and quantities of emergency resources required for the flood accident emergency response based on the state of the flood accident; Step S4: Apply the contents of the first emergency resource list to the corresponding rainfall scenario, and use the cellular automaton model for pre-deduction. According to the pre-deduction results, generate the second emergency resource list, that is, obtain the emergency rescue and dispatch command decision. Among them, the first emergency resource list is pre-deduced using the cellular automaton model to verify the first emergency resources. If the verification directly meets the requirements, the first emergency resource list is converted into the second emergency resource list. If it does not meet the requirements, it is necessary to re-acquire the flood-related information, and combine the deduction results to adjust the emergency resource type and quantity information of the flood accident emergency response. Use the cellular automaton model for secondary deduction to make the emergency resource list meet the requirements, thereby generating the second emergency resource list.

2. The method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario according to claim 1 is characterized in that: In step S2.1: Validation as a key risk factor for input data , whether it satisfies If the data format meets the requirements, use the formula Perform normalization processing. If the requirements are not met, perform data verification and re-obtain input data for judgment. represents a sequence of frames of time-tagged input data, 、 They are all set letters, Represents the input data before normalization; The loss evaluation model in step 2.2 is expressed as: in, express At time , the description value of a certain disaster variable of underground space flood, express At time , the description value of a certain carrier variable of the underground station flood, express At time , the description value of a certain control variable of the underground station flood, The format of the input data for the loss assessment model is represented by Indicates the format of the output data of the loss assessment model, Represents the normalized input data; In step 2.4, the risk population includes casualties on site, missing persons, urgently relocated and resettled persons, and trapped persons. The risk population evaluation criteria include: if the total number of risk population is greater than or equal to 100, the corresponding risk population evaluation value is 5 points; if the total number of risk population is greater than or equal to 50 but less than 100, the corresponding risk population evaluation value is 4 points; if the total number of risk population is greater than or equal to 10 but less than 50, the corresponding risk population evaluation value is 3 points; if the total number of risk population is greater than or equal to 3 but less than 10, the corresponding risk population evaluation value is 2 points; if the total number of risk population is less than 3, the corresponding risk population evaluation value is 1 point; The total economic loss includes the economic losses of construction site equipment loss, construction period loss and material loss at the construction site after the flood. The economic loss evaluation standards include that the total economic loss is greater than or equal to 1 million yuan, and the corresponding economic loss evaluation score is 5 points, indicating that particularly serious economic losses are caused; the total economic loss is greater than or equal to 500,000 yuan and less than 1 million yuan, and the corresponding economic loss evaluation score is 4 points, indicating that serious economic losses are caused; the total economic loss is greater than or equal to 100,000 yuan and less than 500,000 yuan, and the corresponding economic loss evaluation score is 3 points, indicating that relatively heavy economic losses are caused; the total economic loss is greater than or equal to 10,000 yuan and less than 100,000 yuan, and the corresponding economic loss evaluation score is 2 points, indicating that certain economic losses are caused; the total economic loss is less than 10,000 yuan, and the corresponding economic loss evaluation score is 1 point, indicating that basically no economic losses are caused.

3. The method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario according to claim 1 is characterized in that: In step S2 The overall indicator is the risk of underground space flooding under heavy rain scenarios. The first-level indicators include the hazard of hazard factors, sensitivity of the disaster-prone environment, vulnerability of the disaster-bearing body, and disaster prevention and mitigation capabilities. Second-level indicators include the 24-hour maximum rainfall, rainfall duration, construction management and responsibility implementation for the hazard factor risk; the catchment area, drainage network density, surrounding environmental complexity, and rivers for the sensitivity of the disaster-prone environment; the drainage ditch capacity and foundation pit depth for the vulnerability of the disaster-bearing body; and the height of the retaining wall, disaster monitoring and early warning technology, and emergency material reserves for disaster prevention and mitigation capabilities. The whitening weight function for the quantitative risk assessment of rainstorm floods around underground spaces is established, namely the grey clustering evaluation method. The risk level is divided into five grey classes based on the whitening weight function and the quantitative risk characteristics of rainstorm floods around underground spaces. The reference threshold and corresponding risk level of each gray class are obtained, and the first gray class is assigned based on the reference threshold and risk level of each gray class. The weights of all secondary indicators under the first-level indicator ,Among them, the quantitative risk characteristics of rainstorm flooding around underground space are the secondary indicators in the ,multi-index comprehensive evaluation system. When the gray category is 1, the ,reference threshold is [0, 9, ∞], and the risk level is very safe. When the gray category is 2, ,the reference threshold is [0, 7, 14], and the risk level is safe. When the gray category is 3, ,the reference threshold is [0, 5, 10], and the risk level is general. When the gray category is 4, ,the reference threshold is [0, 3, 6], and the risk level is dangerous. When the gray category is 5, ,the reference threshold is [0, 1, 2], and the risk level is very dangerous. The formula of the whitening weight function is: Where, Indicates expert In the Under the first-level indicators Scoring of the secondary indicators, , 、 、 、 、 Respectively represent First-level indicators The risk level of each secondary indicator score under each gray category; Construct a secondary grey evaluation matrix. The specific steps are as follows: Depend on Find the first First-level indicators Secondary indicators The evaluation coefficient in each gray class is as follows: Calculate the secondary indicators separately The evaluation coefficient in the five gray categories is as follows: in, Indicates the Gray class, represents the total clustering coefficient, ; Based on the overall clustering coefficient and evaluation coefficient Get the gray evaluation weight value: Arrange the gray categories The grey evaluation weights of all secondary indicators under the first-level indicators are used to establish a secondary grey evaluation matrix : 。 4. The method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm situation according to claim 1 is characterized in that: The specific steps of step S4 are: Step S4.1: Establish a hierarchical emergency decision-making mechanism based on leadership decision-making, professional handling, and public response; Step S4.2: Based on the professional handling in the emergency decision-making mechanism and the first emergency rescue resource list, the contents of the first emergency resource list are applied to the corresponding rainfall scenario, and a cellular automaton model is used for pre-simulation. According to the pre-simulation results, a second emergency resource list is generated, that is, an emergency rescue and dispatch command decision is obtained, wherein the emergency rescue and dispatch decision includes generating personnel evacuation paths and generating material utilization directions.

5. The method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario according to claim 4 is characterized in that: In the step S4.2, The specific steps of generating a disaster evacuation path for personnel are: Step 4.21: Obtain basic information about the underground space and construct a digital model of the underground space using 3D modeling software. The basic information includes CAD drawings of the underground space, 3D scanning data, camera distribution, and access control system information. Step 4.22: Conduct disaster simulation based on the pre-simulation results of the underground space to determine all possible evacuation paths in the underground space under the current scenario; Step 4.23: Determine the optimal evacuation path based on the determined underground evacuation paths under the current situation and the Dijkstra algorithm. The specific steps are: Step 4.

231. Create a directed weighted graph G(V, E, W), where V is used to store all nodes in the underground space disaster avoidance path under the current scenario, E is used to store the edges connecting two nodes in the underground space disaster avoidance path under the current scenario, and W is used to store all navigable paths in the underground space disaster avoidance path under the current scenario, i.e., the disaster avoidance path. Step 4.232: Define a set D to store the emergency evacuation nodes in the underground space disaster avoidance path under the current situation; Step 4.

233. Create a set P to store the sequence of all nodes in the directed weighted graph G(V,E,W). Create a set Q to record all node paths connecting each node in the directed weighted graph G(V,E,W). Step 4.234: Based on each emergency evacuation node in set D, use the Dijkstar algorithm to perform path resolution on sets P and Q, obtain a shortest path, and simultaneously construct sets M and N. The points and edges on the shortest path are traversed and stored in sets M and N, respectively. Step 4.235: For each emergency evacuation node, remove the edge of the shortest path stored in the set N in step 4.234 from the corresponding directed weighted graph G, so that the directed weighted graph G becomes a directed weighted graph G1; Step 4.236: If there is no traversable path from each emergency evacuation node to the target node in the directed weighted graph G1, proceed to step 42.

37. Otherwise, proceed to step 4.233 and continue intermittently based on the directed weighted graph G1 until all shortest paths from each emergency evacuation node to the target node or the first z shortest paths are obtained, and the loop ends. If all shortest paths are less than z, all shortest paths from each emergency evacuation node to the target node are obtained. Otherwise, the first z shortest paths are obtained. Step 42.37: Given a threshold target, if all shortest paths or the shortest paths of the first z paths are less than or equal to the threshold target, all shortest paths are considered the optimal disaster avoidance paths. Otherwise, a screening function is used to determine the optimal disaster avoidance paths with a number of paths equal to the threshold target among all shortest paths or the shortest paths of the first z paths. Step 4.24: During the disaster avoidance process, re-execute step 4.21 based on the real-time monitoring of rainfall conditions and changes in the underground space; The material utilization direction is generated by using a CNN neural network model and pre-deduction results. The CNN neural network model includes two objective functions and five constraints that need to be met. Objective function 1 is used to represent the lowest cost of emergency material dispatch: ； in, Indicates the number of points for emergency supplies rescue. Indicates the number of disaster-affected points. Indicates that from Order emergency supplies to rescue points To the Disaster-affected areas Dispatch emergency supplies the number of Indicates that from Order emergency supplies to rescue points To the Disaster-affected areas Dispatch Number of times the material is planted; Objective function 2 is used to express the shortest adjustment time of emergency supplies; ； in, Indicates that the Order emergency supplies to rescue points To the Disaster-affected areas Triangular fuzzy number of dispatching time for allocating emergency supplies, , is the pessimistic value of the fuzzy number, is the normal value of the fuzzy number, is the optimistic value of the fuzzy number, Indicates that from Order emergency supplies to rescue points Whether to deploy materials to Disaster-affected areas , ; There are five constraints: Constraint 1 indicates that the actual total supply of emergency supplies at each emergency rescue point is equal to the total demand for emergency supplies at each disaster site. The formula is: in, Indicates the Order emergency supplies to rescue points No. Kinds of materials The actual supply of Indicates the Disaster-stricken areas with emergency supplies The actual demand; Constraint 2 indicates the actual dispatch of emergency supplies from each rescue point to the The first disaster site emergency supplies The amount is equal to the emergency supplies of the disaster site The demand is: in, Indicates the The emergency supplies were dispatched to the rescue point. Disaster-stricken areas with emergency supplies the number of Indicates the The demand for emergency supplies at the disaster-stricken sites; Constraint 3 means that Emergency supplies are dispatched to various disaster-stricken areas. The sum of the quantity of emergency supplies is equal to the amount of emergency supplies transported to the rescue point The actual supply is: Constraint 4 means that Order emergency supplies to rescue points Deploy to Disaster-affected areas The quantity of emergency supplies is a non-negative number, and the formula is: Constraint 5 means that Order emergency supplies to rescue points Whether to deploy emergency supplies to Disaster-affected areas If there are emergency supplies, Order emergency supplies to rescue points Deploy to Disaster-affected areas ,but The value of is 1, otherwise it is 0; in, Indicates the number of types of substances.

6. The method for constructing an underground space flood emergency rescue and dispatch command model under a rainstorm scenario according to claim 1, characterized in that: The cellular automaton model simulation in steps S1 and S4 is based on the rainstorm scenario data, the underground space construction model and the catchment area, using the cellular automaton model to perform physical simulation and numerical simulation of the disaster evolution to obtain disaster data, including the underground space flooded area, flooded water depth and flood arrival time; Among them, the underground space construction model is obtained by obtaining underground space data and importing it based on the underground space data after remodeling according to the drawings. The underground space data includes DEM data, terrain model, remote sensing image, BIM model, rainstorm data and meteorological data. The terrain model refers to the underground space model. The rainstorm scenario data and the catchment area are obtained based on the rainfall parameter information in the underground space data and the obtained rainstorm intensity.

7. A device for constructing an underground space flood emergency rescue and dispatch command model under heavy rain scenarios, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 6.

Images